J . Am. Chem. SOC.1992, 114, 7919-7920 is inferior to the Norit-A/silica gel because, with the former, packing is more difficult and cracking of the stationary phase occurs easily. Molecular sieves (13X pellets ground to a fine powder and packed in a column) were used as a stationary phase and proved to be inadequate for c 6 0 purifications. Acknowledgment. This research was funded by the Department of the Navy, Office of the Chief of Naval Research, Young Investigator Award Program (1989-1992), the National Science Foundation (RII-8922165, DMR-9158315), and generous industrial contributors to the NSF Presidential Young Investigator Award (1991-1996) for J.M.T.: Hercules Incorporated, IBM Corporation, Ethyl Corporation, and the Shell Development Company. W.A.S. thanks NASA and the American Vacuum Society for scholarships. We thank Dr. Nick Griffith of the Aluminum Company of America for providing us with alumina for the comparison studies.
7919
Scheme I
-
Me‘CU(CN)LI,
1,4-adduct
u
Scheme I1
w
mR
j
R
Q -
---
3e
k‘ > aoR
(a) R OEt (b)R OMe
IC) R N(iPr)?
Q
(d) R N(TMS)z (e) R = OSI(kPr)3
Preparation and Use of Vinylic Lithio cyanocuprates Containing an w-Electrophilic Center Bruce H. Lipshutz* and Robert Keil
Table I. Formation and Reactions of Functionalized Lithiocuprates Entry
1-Alkyne
Enone
Product‘
0
Department of Chemistry, University of California Santa Barbara, California 931 06 Received November 21, 1991
Traditional cuprate formation, as originally prescribed by Gilman,’ bears a fundamental limitation in its reliance on organolithium precursors (Le., 2RLi + CuX). Thus, lithiocuprates of either the lower order (LO, R2CuLi) or higher order (HO, R2Cu(CN)Li2)persuasion which contain electrophilic centers (e.g., an ester or nitrile group) in R are presently unknown.2 To circumvent the incompatibility of a highly reactive organolithium containing such a useful functionality, more stable organometallics have been developed, most notably from the K n ~ h e l Rieke: ,~ and Piers2 groups. However, the price paid for switching from lithium to other gegenionsj is the lowering of cuprate reactivity, a general observation characteristic of both neutral species (Le., “RCU.L~X”)~ as well as ate complexes (R2CuM6a/R2Cu(CN)MM’6b). In this report, we now describe the first method for generation of vinylic lithiocuprates which contain internal electrophiles utilizing a transmetalation strategy based on readily available zirconium intermediates 1’ (Scheme I). Treatment of a vinylzirconate 1, easily prepared from 1-alkynes using Cp2Zr(H)C1 in T H F (room temperature, 15 min),8 with the trivial H O cuprate Me2Cu(CN)Li2at low temperatures (-78 OC, 15 min) leads directly to the mixed cuprate 2. Introduction of an a,&unsaturated ketone, neat or in THF, to the newly generated cuprate at this temperature affords the expected 1,4( 1 ) Gilman, H.; Jones, R. G.; Woods, L. A. J. Org. Chem. 1952,17, 1630. (2) There is an example of an LO lithiocuprate containing a primary chloride, see: Piers, E.; Roberge, J. Y . Tetrahedron Leu. 1991, 32, 5219. Piers, E.; Yeung, B. W. A. J . Org. Chem. 1984, 49, 4567. Piers, E.; Karunaratne, V. J . Chem. Soc., Chem. Commun. 1983, 935. (3) Rao, S. A.; Knochel, P. J . Am. Chem. SOC.1991, 113, 5735 and references therein. (4) Stack, D. E.; Dawson, B. T.; Rieke, R. D. J . A m . Chem. SOC.1991, 113, 4672 and references therein. See also: Ebert, G. W.; Klein, W. R. J . Org. Chem. 1991, 56,4144. (5) Neutral complexes of this type are normally relatively unreactiveSb unless used in the presence of, for example, a Lewis acid, see: Yamamoto, Y. Angew. Chem., Inr. Ed. Engl. 1986, 25, 947. (b) Lipshutz, B. H.; Sengupta, S. Org. React., in press. Posner, G. H . Ibid. 1975, 22, 253; Ibid. 1972, 19, 1 . (6) (a) Bertz, S. H.; Gibson, C . P.; Dabbagh, G. Organometallics 1988. 7, 227. (b) Lipshutz, B. H.; Ellsworth, E. L.; Behling, J. R.; Campbell, A. L. Tetrahedron Lett. 1988,29,893. Lipshutz, B. H.; Nguyen, S. L.; Parker, D. A.; McCarthy, K. E.; Barton, J.; Whitney, S.; Kotsuki, H. Tetrahedron 1986, 42, 2873. (7) Negishi, E. Chem. Scripra 1989, 29, 457. Schwartz, J.; Labinger, J . A. Angew. Chem., Inr. Ed. Engl. 1976, IS, 333. Negishi, E. Pure Appl. Chem. 1981,53,2333. Negishi, E.; Takahashi, T. Aldrichimica Acta 1985, 18, 31; Synthesis 1988, 1 . (8) Lipshutz, B. H.; Ellsworth, E. L. J . Am. Chem. Soc. 1990, 11 2, 7440.
0002-7863/92/1514-79 19$03.00/0
Yleld(%)b
0
-cN
L
0
C
N
75
0
7lC’@
7
0
0
Fully characterized by IR, N M R , M S , and H R M S data. Isolated yields. cBF,.Et20 (1 equiv) was added to the cuprate prior to introisomer by capillary GC. CYieldwas 35% duction of the enone. without BF,.Et20.
adduct in good isolated yields. This simple, one-pot process can be applied to acetylenes which possess a nitrile, ester, or chloride residue (Table I). It is especially noteworthy that &&disubstituted enones react readily at -78 OC, a clearly distinguishable feature between these lithiocuprates and, for example, zinc halide-containing reagent^.^ Moreover, the rapidity and simplicity associated with this hydrozirconation-transmetalation-Michael addition sequence are also attractive elements, as there is no major time commitment to prior generation of activated organometallics (Le., RZnX2 or CU(O)~). An unexpected observation was made in the case of 5-hexynoic acid ethyl or methyl ester (3a,b), where the sequence described (9) Knoess, H . P.; Furlong, M. T.; Rozema, M. J.; Knochel, P. J . Org. Chem. 1991, 56, 5974. Knochel, P.; Chou, T.-S.; Chen, H. G.; Yeh, M. C. P.; Rozema, M . J . Ibid. 1989, 54, 5202. Knochel, P.; Yeh, M. C. P.; Bcrk, S. C.; Talbert, J . Ibid. 1988, 53, 2390.
0 1992 American Chemical Society
J . Am. Chem. SOC.1992, 114, 7920-7922
7920 Scheme 111 C P ? r/c' WFG
Me
+
Me2Cu(CN)L12
-
u
above led to only olefinic products 4 (R = OEt, OMe) (Scheme 11). These failures to transmetalate may be a consequence of intra- and/or intermolecular chelation of the ester carbonyl with the Zr(1V) present (cf. 6 and 7).1° This notion stems from the
Acknowledgment. Financial support provided by the NIH (GM 40287) and the donors of the Petroleum Research Fund, administered by the American Chemical Society, is gratefully acknowledged. We are indebted to Dr. Toshio Tanaka (Teijin Ltd.) for a generous sample of the nonracemic cyclopentenone used in Table I, entry 8. Registry No. 1, 139461-23-7; 38, 108545-38-6; 3b, 77758-51-1; 3c, 139461-32-8; 3d, 139461-33-9; 3 (R = OSiMe,Bu-r), 139461-36-2; 48, 54653-25-7; 4b, 2396-80-7; 4c, 139461-34-0; 4, 139461-35-1; Cp,Zr(H)CI, 37342-97-5; Me2Cu(CN)Li2, 80473-70-7; HC=C(CH2),CN, 14918-21-9; H e C ( C H 2 ) 2 0 C O P h , 122471-85-6; HC&(CH,),CI, 5 1908-64-6; 2-cyclohexenone, 930-68-7; 2-methyl-2-cyclopentenone, 1 120-73-6; 4-methylpent-3-en-2-one,107-86-8; 3-methyl-2-cyclopentenone, 1 193- 18-6; (R)-4-(dimethyl-rert-butylsiloxy)-2-cyclopentenone, 61305-35-9; (E)-3-(5-cyano-l-pentenyl)cyclohexanone, 13946 1-24-8; (E)-3-[6-(triisopropylsiloxy)-6-0~01-hexanyl]cyclohexanone, 139461-25-9; (E)-3- [4-(phenylcarbonyloxy)-1-butenyl]cyclohexanone, 139461-26-0; (E)-3-(4-chloro-l-butenyl)cyclohexanone, 139461-27-1; 2-methyl-3- [6-(triisopropylsiloxy)-6-0~0-1-hexenyl]cyclopentanone, 139461-28-2; triisopropylsilyl (E)-7,7-dimethyl-9-oxodec-5enoate, 13946 1-29-3; (E)-3-methyl-3- [6-(triisopropylsiloxy)-6-0~0-1 hexenyl]cyclohexanone, 139461-30-6; (E)-2(R)-(dimethyl-rert-butylsiloxy)- 3 (3)-[ 6 4 triisopropylsiloxy)-6-0~0-1 -hexenyl]cyclopentanone, 139461-3 1-7.
-
Q
Q
observation that for all known HO cyanocuprate-based transmetalations (involving Sn," Te,I2 Al,I3 Zr8J3), whatever the mechanism involved, the common denominator is the Lewis acidic nature of the participating organometallic. Thus, occupation of the remaining coordination site as in 617 shuts down ligand exchange with Me2Cu(CN)Li2and, hence, vinylcuprate formation. However, replacement of R = OEt or OMe in 3a,b with R = OSi(i-Pr)3 (OTIPS), on sterids or stereoelectronic grounds (or both),I6 completely restores the transmetalation pathway. In support of the argument above, we also find that (1) introduction of ethyl heptanoate to reactions of 3e completely inhibits formation of 5 (R = OTIPS) and (2) all attempts to effect the sequence with amides 3c,d likewise produced the alkene 4 rather than the desired Michael adducts 5. The true nature of the species formed in the transmetalation process has yet to be established. Although formulated here initially as a HO cuprate (cf. Scheme I), it is possible that 2 reacts to some degree with the presumed Cp2Zr(Me)C1 byproduct of the transmetalation to produce an LO lithiocyanocuprate 8 (Scheme 111). Spectroscopic studies are anticipated to elucidate this issue. In summary, the extreme mildness and rapidity with which vinylzirconates undergo transmetalations with HO cyanocuprates have led to the first examples of functionalized lithioc~prates.'~ Applications, in particular to polyene macrolide chemistry, are currently in progress and will be reported in due course. (IO) See, for examples: Yamamoto, Y.; Komatsu, T.; Maruyama, K. J . Organomef.Chem. 1985, 285, 31. Buchwald, S.L.; Nielsen, R. B.; Dawan, J. C. Organomerallics 1988, 7, 2324. ( 1 1) For example, see Behling, J. R.;Babiak, K. A.; Ng, J. S.;Campbell, A. L.; Moretti, R.; Koerner, M.; Lipshutz, B. H. J . Am. Chem. SOC.1988, I 12, 7440. (12) Comasseto, J. V.; Berriel, J. N. Synrh. Commun. 1990, 20, 1681. (13) Ireland, R. E.; Wipf, P.J . Org. Chem. 1990, 55, 1425. (14) Babiak, K. A,; Behling, J. R.;Dygos, J. H.; McLaughlin, K. T.; Ng, J. S.; Kalish, V. J.; Kramer, S. W.; Shone, R. L. J . Am. Chem. SOC.1990, 112, 7441. (15) Use of the corresponding r-BuMe2Siester gave low (15-209) yields. (16) Frye, S. V.; Eliel, E. L. Tetrahedron Lerr. 1986, 27, 3223; J . Am. Chem. SOC.1988, 110, 484. Chen, X.;Hortelano, E. R.; Eliel, E. L.; Frye, S . V. Ibid. 1990,112,6130. Magnus, P.; Mugrage, B. Ibid. 1990, 112,462. (17) A typical procedure for the preparation of an w-functionalized cyanocuprate is as follows: The vinyl zirconate is formed by treating a slurry of Cp2Zr(H)CI (0.5 mmol) in anhydrous THF (2 mL) with the I-alkyne (0.5 mmol, added neat) for 15 min at room temperature. The solution is then cooled to -78 OC and treated for IO min with a -78 OC THF solution of Mc2Cu(CN)Li2(prepared from 2MeLi (1 mmol) and CuCN (0.5 mmol) at -78 'C in 2 mL of THF). If necessary, BF3.Et20(0.5 mmol in 1 mL of EtzO or THF) can be added at this time in a dropwise fashion. The enone (0.25 mmol, neat or in 0.5 mL of THF) is then introduced and the solution stirred at -78 'C until the reaction is complete (5 min to 1 h), as determined by TLC. The mixture is then quenched by pouring it into saturated aqueous NH,Cl/concentrated NH,OH solution (9: 1) and is extracted with EtzO. Normal handling and chromatography affords the pure 1,4-addition product.
Supplementary Material Available: Listings of spectral data for the products in Table I and IH and 13CNMR spectra for these and other products of the reactions of lithiocuprates (18 pages). Ordering information is given on any current masthead page.
Dynamics of Solute Motion: Photoisomerization Shows Linear Dependence on Solvent Mass Bjiirn Sauerwein, Sean Murphy, and Gary B. Schuster* Roger Adams Laboratory, Department of Chemistry University of Illinois, Urbana, Illinois 61801 Received May 28, I992
In the general theory of solute motion based on Kramers' equation,I the control of reaction dynamics is divided into two limiting categories. At low friction (i.e., at low viscosity), the reaction rate is expected to increase with the number of collisions with solvent, as these supply the energy required to cross a potential barrier. At high viscosity, solvent will obstruct the path along the reaction coordinate and decrease the rate. Experiments have shown that Kramers' theory regularly fails when the potential barrier of the reaction is small (relative to kT) or no activation energy is requireda2 Several attempts have been made to account for behavior in this low-barrier regime.3 Grote and Hynes4 replaced the friction coefficient in Kramers' equation (E), with a frequency-dependent function, [(a), which permits the viscosity to vary with the rate of the reaction. Velsko and Fleming2 added a reaction coordinate-dependent sink function to Kramers' theory and obtained an improv$ fit to experimental observation. In an approach developed by Akesson, Sundstrom, and G i l l b r ~the ,~ barrier height, large or small, is made to depend on the solvent. When this approach is applied to a homologous solvent series such as the n-alcohols, adjusting the barrier generates a good fit to experimental data. Dote, Kivelson, and Schwa& introduced "free spaces" in the solvent where rotation can take place unhindered ( I ) Kramers, H. A. Physica 1940, 7 , 284. (2) Velsko, S. P.; Waldeck, D. H.; Fleming, G. R.J. Chem. Phys. 1983,
78, 249. (3) Barbara, P.; Walker, G. C. Reu. Chem. Interm. 1988, 10, 1. Bagchi, B.; Fleming, G. R. J . Phys. Chem. 1990, 94, 9. (4) Grote, R. F.; Hynes, J. T. J . Chem. Phys. 1980, 73, 2715. Grote, R. F.; Hynes, J. T. J. Chem. Phys. 1981, 74, 4465. (5) Akesson, E.; Sundstrom, V.; Gillbro, T. Chem. Phys. 1986,106,269. (6) Dote, J. L.; Kivelson, D.; Schwartz, R. N. J . Phys. Chem. 1981, 85, 2169.
0002-7863/92/1514-7920$03.00/00 1992 American Chemical Society